NORTHCUTT BIKES: THE FORECASTING PROBLEM
Jan Northcutt, owner of Northcutt Bikes, started business in 1995 when she noticed
the quality of bikes she purchased for sale in her bike shop declining while the prices
went up. She also found it more difficult to obtain the features she wanted on ordered
bikes without waiting for months. Her frustration turned to a determination to build
her own bikes to her particular customer specification.
She began by buying all the necessary parts (frames, seats, tires, etc.) and
assembling them in a rented garage using two helpers. As the word spread about
her shop's responsiveness to options, delivery, and quality, however, the individual
customer base grew to include other bike shops in the area. As her business grew
and demanded more of her attention, she soon found it necessary to sell the bike
shop itself and concentrate on the production of bikes from a fairly large leased
factory space.
As the business continued to grow, she backward integrated more and more
processes into her operation, so that now she purchases less than 50% of the component
value of the manufactured bikes. This not only improves her control of production
quality but also helps her control the costs of production and makes the final
product more cost attractive to her customers.
The Current Situation
Jan considers herself a hands-on manager and has typically used her intuition and her
knowledge of the market to anticipate production needs. Since one of her founding
principles was rapid and reliable delivery to customer specification, she felt she
needed to begin production of the basic parts for each particular style of bike well in
advance of demand. In that way she could have the basic frame, wheels, and standard
accessories started in production prior to the recognition of actual demand, leaving
only the optional add-ons to assemble once the order came in. Her turnaround time
for an order of less than half the industry average is considered a major strategic
advantage, and she feels it is vital for her to maintain or even improve on response
time if she is to maintain her successful operation.
As the customer base has grown, however, the number of customers Jan
knows personally has shrunk significantly as a percentage of the total customer
base for Northcutt Bikes, and many of these new customers are expecting or even
demanding very short response times, as that is what attracted them to Northcutt
Bikes in the first place. This condition, in addition to the volatility of overall
demand, has put a strain on capacity planning. She finds that at times there is a lot
of idle time (adding significantly to costs), whereas at other times the demand
exceeds capacity and hurts customer response time. The production facility has
therefore turned to trying to project demand for certain models and actually building
a finished goods inventory of those models. This has not proven to be toosatisfactory, as it has actually hurt costs and some response times. Reasons include
the following:The finished goods inventory is often not the "right" inventory, meaning shortagesfor some goods and excessive inventory of others. This condition bothhurts responsiveness and increases inventory costs.Often to help maintain responsiveness, inventory is withdrawn from finishedgoods and reworked, adding to product cost.Reworking inventory uses valuable capacity for other customer orders, againresulting in poorer response times and/or increased costs due to expediting.Existing production orders and rework orders are both competing for vitalequipment and resources during times of high demand, and scheduling hasbecome a nightmare.The inventory problem has grown to the point that additional storage space isneeded, and that is a cost that Jan would like to avoid if possible.Another problem Jan faces is the volatility of demand for bikes. Since she isworried about unproductive idle time and yet does not wish to lay off her workersduring times of low demand, she has allowed them to continue to work steadily andbuild finished goods. This makes the problem of building the "right"finished goodseven more important, especially given the tight availability of storage space.
Past Demand
The following shows the monthly demand for one major product line: the standard26-inch 10-speed street bike. Although it is only one of Jan's products, it is representativeof most of the major product lines currently being produced by NorthcuttBikes. If Jan can find a way to use this data to more constructively understand herdemand, she feels she can probably use the same methodologies to project demandfor other major product families. Such knowledge can allow her, she feels, to planmore effectively and continue to be responsive while still controlling costs.
ACTUAL DEMAND
Month 2008 2009 2010 2011
January 437 712 613 701
February 605 732 984 1291
March 722 829 812 1162
April 893 992 1218 1088
May 901 1148 1187 1497
June 1311 1552 1430 1781
July 1055 927 1392 1843
August 975 1284 1481 839
September 822 1118 940 1273
October 893 737 994 912
November 599 983 807 996
December 608 872 527 792
Questions for the Case Study
1. Plot the data and describe what you see. What does it mean and how would you use the information from the plot to help you develop a forecast?
2. Use at least two different methodologies to develop as accurate a forecast forthe demand as possible. Use each of those methods to project the next fourmonths demand.
3. Which method from question 2 is "better"? How do you know that?
4. How, if at all, could we use Jan's knowledge of the market to improve the forecast?Would it be better to forecast in quarterly increments instead of monthly?Why or why not?
5. Are there other possible approaches that might improve Jan's operation andsituation? What would they be and how could they help?
6. Has Jan's operation grown too large for her to control well? Why or why not? What would you suggest she do? What additional information would you suggestshe look for to help her situation?